Gfree Bio proposes to collaborate with Professor Bajaj and his group at the University of Texas to develop a general and much more theoretically rigorous algorithm for fitting the ligand binding region of experimental protein-ligand cocrystal structures. Current methods and practices often lead to very poor geometries for the bound ligand. This is a significant impediment to the most effective use of crystallography in drug discovery. We will solve this problem by developing an automated method for using polar sampling and optimization in conjunction with the experimental electron density information to correct protein-ligand cocrystal structures. This methodology will capitalize on recent advances in the Bajaj lab with respect to fast multidimensional optimization, linear scaling electrostatics and solvation methods, and advanced data structures for representing molecular systems. This software has great commercial potential since the problem of properly fitting bound ligands has been receiving increasing attention in the literature, and industry. As such our solution will have broad impact on structure-based drug design.